INDEX

Explanations

instances of the word "a" followed by various descriptors indicating quantity or type

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

a basis

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

 doua

-0.38

thumb

-0.38

 spit

-0.37

sugges

-0.37

 зас

-0.35

 Wikimedijinoj

-0.35

 جایی

-0.35

ammer

-0.35

wit

-0.34

 tombé

-0.33

POSITIVE LOGITS

échelle

0.61

 échelle

0.60

 scale

0.54

 skala

0.53

 basis

0.52

mergeFrom

0.52

 Scale

0.52

 escala

0.51

Scale

0.48

editForm

0.48

Activations Density 0.024%

instances of the word "a" followed by various descriptors indicating quantity or type

a basis

No Comments

No Known Activations

instances of the word "a" followed by various descriptors indicating quantity or type

a basis

No Comments

No Known Activations