INDEX

Explanations

expressions related to listening or receiving information

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

hearing opinions

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

把它

-0.39

tanleria

-0.37

untur

-0.36

Tig

-0.35

waltung

-0.34

 fizz

-0.34

 Mainland

-0.34

 Matrix

-0.33

 matrix

-0.33

 supplier

-0.33

POSITIVE LOGITS

 hear

0.81

Hear

0.66

hear

0.63

 hears

0.62

 Hear

0.61

 hearing

0.61

 HEAR

0.61

 heard

0.59

Hearing

0.59

 Hearing

0.59

Activations Density 0.004%

expressions related to listening or receiving information

hearing opinions

No Comments

No Known Activations