INDEX

Explanations

mentions of public figures giving speeches or statements in the news

oai_token-act-pair · gpt-3.5-turbo

instances of the word "spoke" indicating dialogue or speech by various individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

aredevil

-0.72

rypt

-0.70

 ILCS

-0.66

mental

-0.66

rawler

-0.66

anny

-0.65

assi

-0.64

brance

-0.64

ween

-0.62

involved

-0.62

POSITIVE LOGITS

 volumes

0.87

 aloud

0.86

 omin

0.77

 transcripts

0.76

 harshly

0.72

 onstage

0.72

 louder

0.70

 passionately

0.70

 glow

0.70

 mute

0.69

Activations Density 0.029%

mentions of public figures giving speeches or statements in the news

instances of the word "spoke" indicating dialogue or speech by various individuals

No Comments

No Known Activations