INDEX

Explanations

numeric expressions in a specified time format

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

time references

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

 esternos

-0.64

Rptr

-0.57

writeFieldEnd

-0.54



-0.54

 Winaray

-0.54

lichter

-0.54

 Verk

-0.53

transQ

-0.53

addPreferredGap

-0.53

criminator

-0.52

POSITIVE LOGITS

 prêtres

0.36

 sacerdotes

0.35

 nocturno

0.34

 médicale

0.34

ApiModelProperty

0.34

 mediodía

0.33

 naciones

0.33

 katholischen

0.33

 personnalisé

0.33

 żyw

0.33

Activations Density 0.000%

numeric expressions in a specified time format

time references

No Comments

No Known Activations