INDEX

Explanations

references to financial issues and personal challenges

oai_token-act-pair · gpt-4o-mini Triggered by @bot

masturbating, she, Beijing

np_acts-logits-general · gemini-2.5-flash-lite

the neuron detects salient factual/content-bearing tokens — i.e., important words like named entities, dates, numbers, and other prominent topic words in news-like text.

oai_token-act-pair · gpt-5-mini Triggered by @vetterc0

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

rungsseite

-0.78

 &___

-0.69

 للمعارف

-0.67

:✨

-0.60

ThroughAttribute

-0.56

elemField

-0.55

Дереккөздер

-0.53

delwed

-0.52

 INTERESAR

-0.52

-------------</

-0.52

POSITIVE LOGITS

ayuno

0.40

 Sincerely

0.39

stützung

0.39

Cite

0.36

Sincerely

0.36

zem

0.34

Aid

0.33

 dock

0.32

dă

0.32

thest

0.32

Activations Density 0.067%

references to financial issues and personal challenges

masturbating, she, Beijing

the neuron detects salient factual/content-bearing tokens — i.e., important words like named entities, dates, numbers, and other prominent topic words in news-like text.

No Comments

No Known Activations

references to financial issues and personal challenges

masturbating, she, Beijing

the neuron detects salient factual/content-bearing tokens — i.e., important words like named entities, dates, numbers, and other prominent topic words in news-like text.

No Comments

No Known Activations