INDEX

Explanations

instances of legal terminology and phrases related to legal proceedings

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

Further

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

awtextra

-0.69

+#+#

-0.61

 linkovi

-0.59

ValueStyle

-0.59

AndEndTag

-0.59

AccessorTable

-0.57

ContentAlignment

-0.56

mergeFrom

-0.56



-0.55

RTHOOK

-0.54

POSITIVE LOGITS

 berücksich

0.34

Dep

0.34

 Belastung

0.34

Erreferentziak

0.33

ésult

0.33

Default

0.33

bebasan

0.32

이고

0.32

 nobody

0.32

嘛

0.32

Activations Density 0.143%

instances of legal terminology and phrases related to legal proceedings

Further

No Comments

No Known Activations