INDEX

Explanations

elements related to restrictions or regulations

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

are allowed

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

 HttpNotFound

-0.69

 betweenstory

-0.61

 pleaſure

-0.60

 incrí

-0.60

 ModelExpression

-0.57

 ſind

-0.55

 indígen

-0.54

Personendaten

-0.54

CloseOperation

-0.54

iſen

-0.53

POSITIVE LOGITS

Rhestr

0.35

im

0.32

[*]

0.32

entary

0.32

im

0.32

 beginnetje

0.32

 Biôgrafia

0.31

 же

0.29

 Primary

0.29

 يتيمه

0.29

Activations Density 0.077%

elements related to restrictions or regulations

are allowed

No Comments

No Known Activations

elements related to restrictions or regulations

are allowed

No Comments

No Known Activations