INDEX

Explanations

proper nouns, particularly names of individuals and organizations involved in political news

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Configuration

jbloom/Gemma-2b-IT-Residual-Stream-SAEs/gemma_2b_it_blocks.12.hook_resid_post_16384

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

HuggingFaceFW/fineweb

Features

16,384

Data Type

float32

Hook Name

blocks.12.hook_resid_post

Hook Layer

Architecture

standard

Context Size

1,024

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Coefficients

-0.52

 burton

-0.49

 csal

-0.47

 ALFRED

-0.46

 Rossetti

-0.46

 chong

-0.45

 WATSON

-0.44

 bandai

-0.44

 HONORABLE

-0.44

 mutagen

-0.43

POSITIVE LOGITS

Sen

1.35

Sen

1.32

 senator

1.31

 Senator

1.31

sen

1.24

Senator

1.23

 senators

1.20

 Senators

1.13

 Senate

1.04

SEN

1.02

Activations Density 0.042%

proper nouns, particularly names of individuals and organizations involved in political news

No Comments

No Known Activations