INDEX

Explanations

references to "opening" events or statements in various contexts

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

opening of events

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

 Vereinigte

-0.42

 pernas

-0.41

 république

-0.39

électricité

-0.38

setVerticalGroup

-0.37

nalpot

-0.36

 sitä

-0.35

 them

-0.35

 Wissenschaft

-0.35

courir

-0.35

POSITIVE LOGITS

 Opening

0.93

 opening

0.93

opening

0.83

Opening

0.81

 OPENING

0.78

CloseOperation

0.75

ductory

0.74

OPENING

0.73

 initial

0.72

 openings

0.71

Activations Density 0.009%

references to "opening" events or statements in various contexts

opening of events

No Comments

No Known Activations