INDEX

Explanations

references to changes or modifications in data or methods

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

changing words or fixing typos

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

anwal

-0.38

 चला

-0.36

 ostavi

-0.36

InputBorder

-0.34

RRect

-0.33

 empaque

-0.32

lam

-0.32

appartamento

-0.32

intervention

-0.32

tanleria

-0.32

POSITIVE LOGITS

 propOrder

0.66

synthesize

0.50

__(/*!

0.50

 HasFactory

0.47

tonode

0.47

 PeEnEo

0.46

transQ

0.46

 AssemblyProduct

0.45

 fehl

0.44

 beginnetje

0.43

Activations Density 0.059%

references to changes or modifications in data or methods

changing words or fixing typos

No Comments

No Known Activations