INDEX

Explanations

negative sentiments and references to misleading or deceptive practices

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

warn about expectations

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

OGND

-0.67

Vapour

-0.47

PRNewswire

-0.42

 surla

-0.41

tiles

-0.40

opic

-0.38

proof

-0.38

\{\\

-0.37

Dent

-0.37

Hochspringen

-0.37

POSITIVE LOGITS

oa̍t

0.47

 souffrance

0.47

 advirtió

0.47

 prévention

0.44

makeConstraints

0.44

 sufrimiento

0.44

pulsera

0.42

 mobilité

0.41

 sorpresas

0.41

 advir

0.39

Activations Density 0.042%

negative sentiments and references to misleading or deceptive practices

warn about expectations

No Comments

No Known Activations