INDEX

Explanations

phrases and conjunctions indicating relationships between concepts or items

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

sentence endings and new sentences

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

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

 INTELL

-0.41

 Transfer

-0.40

 Intelligence

-0.39

hitheater

-0.39

 initial

-0.38

 nomin

-0.38

 Hygiene

-0.37

 intelligence

-0.37

 initially

-0.37

 intellig

-0.36

POSITIVE LOGITS

 ब्रेकडाउन

0.56

RenderAtEndOf

0.56

 mijne

0.54

 popoli

0.52

 Italijani

0.51

Portály

0.50

cartão

0.50

AutoScaleMode

0.49

SequentialGroup

0.48

RTCK

0.48

Activations Density 0.054%

phrases and conjunctions indicating relationships between concepts or items

sentence endings and new sentences

No Comments

No Known Activations