INDEX

Explanations

references to specific time periods or durations

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

period of time

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

 científicas

-0.46

 Inscrivez

-0.44

 scientifiques

-0.44

smart

-0.43

 Cientí

-0.43

blå

-0.43

black

-0.43

 wissenschaft

-0.42

 Sociales

-0.41

steering

-0.41

POSITIVE LOGITS

 period

1.33

 periods

1.32

 Period

1.31

period

1.24

 Periods

1.23

 PERIOD

1.18

Period

1.16

periods

1.13

PERIOD

1.06

Periods

1.04

Activations Density 0.019%

references to specific time periods or durations

period of time

No Comments

No Known Activations