INDEX

Explanations

questions being asked or reported by individuals

oai_token-act-pair · gpt-3.5-turbo

instances of the word "asked" in various contexts

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

marine

-0.67

execute

-0.65

rats

-0.65

vas

-0.65

endi

-0.62

LOD

-0.62

agement

-0.62

EStreamFrame

-0.62

Fit

-0.61

our

-0.61

POSITIVE LOGITS

 rhet

1.02

 Questions

0.95

 quizz

0.95

 questions

0.90

ioned

0.85

 asked

0.83

 Asked

0.82

 sarcast

0.74

cles

0.73

 questioned

0.72

Activations Density 0.020%

questions being asked or reported by individuals

instances of the word "asked" in various contexts

No Comments

No Known Activations