INDEX

Explanations

references to sources and their attributes in text

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

the word source

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

tanleria

-0.48

ๆ

-0.44

 renamed

-0.40

 prévue

-0.39

ๆ

-0.39

 dentelle

-0.39

 progress

-0.38

 trains

-0.38

 forfait

-0.38

VersionUID

-0.37

POSITIVE LOGITS

 source

1.19

 sources

1.13

 Sources

1.08

source

1.06

 Source

1.03

sources

1.01

 SOURCE

0.92

Sources

0.91

Source

0.90

 SOURCES

0.90

Activations Density 0.081%

references to sources and their attributes in text

the word source

No Comments

No Known Activations