INDEX

Explanations

references to quantities and comparisons

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

quantifiers and others

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

 للمعارف

-0.68

ſelves

-0.61

AddTagHelper

-0.60

ſelf

-0.60

+#+

-0.58

 Paglinawan

-0.56

featureID

-0.53

SourceChecksum

-0.53

 ""],

-0.51

 ſche

-0.51

POSITIVE LOGITS

 ones

0.73

 others

0.64

 Others

0.63

Others

0.61

others

0.58

 quelli

0.52

 celui

0.51

 ours

0.46

 celles

0.46

 Celui

0.43

Activations Density 0.066%

references to quantities and comparisons

quantifiers and others

No Comments

No Known Activations