INDEX

Explanations

questions being asked

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 @ 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

âĶĢâĶĢ

-0.77

absor

-0.73

âĸ¬

-0.71

Ĥ¬

-0.66

EStreamFrame

-0.65

 Scouting

-0.64

Lago

-0.63

ordinate

-0.62

marine

-0.62

\\\\

-0.61

POSITIVE LOGITS

 rhet

1.28

 questions

1.03

him

0.98

 plaint

0.87

 politely

0.86

 probing

0.84

 them

0.83

 respondents

0.82

naires

0.81

 forgiveness

0.80

Activations Density 0.055%

questions being asked

instances of the word "asked" in various contexts

No Comments

No Known Activations