INDEX

Explanations

names of individuals and organizations

oai_token-act-pair · gpt-3.5-turbo

phrases related to political commentary and references to specific individuals or organizations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 customs

-0.74

wives

-0.65

 wearer

-0.61

 doors

-0.61

 Manufacturer

-0.61

Cele

-0.60

window

-0.59

MSN

-0.59

 proceeding

-0.59

facade

-0.58

POSITIVE LOGITS

 insightful

0.93

blogs

0.93

 Slate

0.92

 blog

0.92

 blogs

0.86

pdf

0.86

 aptly

0.85

published

0.84

 essay

0.84

 eloqu

0.83

Activations Density 0.539%

names of individuals and organizations

phrases related to political commentary and references to specific individuals or organizations

No Comments

No Known Activations