INDEX

Explanations

references to corruption in various contexts

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

political corruption

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.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

 gawas

-0.41

irl

-0.38

übers

-0.36

oneofs

-0.35

bbs

-0.35

bs

-0.34

SourceChecksum

-0.33

gesehen

-0.32

ide

-0.32

 Geraadpleegd

-0.32

POSITIVE LOGITS

 Corruption

0.73

 corrom

0.72

 corruption

0.70

 corrupt

0.65

Corruption

0.64

 corrupción

0.63

Scénario

0.63

 corrupted

0.61

corruption

0.60

存于互联网档案馆

0.59

Activations Density 0.003%

references to corruption in various contexts

political corruption

No Comments

No Known Activations