INDEX

Explanations

mentions of speech-related terms

oai_token-act-pair · gpt-3.5-turbo

references to speech and its various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Soldiers

-0.78

 Associ

-0.69

xon

-0.63

 Mirage

-0.63

 Leone

-0.63

 Dani

-0.62

Ã±o

-0.61

 Turtles

-0.60

 Mississ

-0.59

 Alto

-0.59

POSITIVE LOGITS

 imped

0.96

writer

0.95

writers

0.91

writing

0.90

speech

0.86

 uttered

0.84

anooga

0.82

 speech

0.81

press

0.79

emic

0.78

Activations Density 0.044%

mentions of speech-related terms

references to speech and its various contexts

No Comments

No Known Activations