INDEX

Explanations

references to behind-the-scenes aspects or insider views of situations

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

behind the scenes

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

 Constitucional

-0.47

 dép

-0.44

 marcador

-0.44

prüche

-0.42

alnız

-0.41

BufferException

-0.40

Abit

-0.40

 predominant

-0.40

 semblable

-0.39

Приятного

-0.38

POSITIVE LOGITS

 backstage

0.98

 behind

0.82

behind

0.82

 Behind

0.79

Behind

0.78

 BEHIND

0.77

 insider

0.71

 detrás

0.69

 INSIDE

0.68

StoryboardSegue

0.64

Activations Density 0.006%

references to behind-the-scenes aspects or insider views of situations

behind the scenes

No Comments

No Known Activations