INDEX

Explanations

references to sources in a given text

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

listing sources

np_acts-logits-general · gemini-2.5-flash-lite

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 this

-0.54

the

-0.43

 Schild

-0.43

-0.42

 This

-0.42

 that

-0.42

emp

-0.40

payment

-0.40

this

-0.40

<bos>

-0.39

POSITIVE LOGITS

 sources

1.81

 Sources

1.76

sources

1.72

Sources

1.68

 SOURCES

1.55

SOURCES

1.30

 fuentes

1.23

 Quellen

1.22

 fontes

1.13

ources

1.12

Activations Density 0.011%

references to sources in a given text

listing sources

No Known Activations