INDEX

Explanations

occurrences related to statements made by individuals to the press

oai_token-act-pair · gpt-3.5-turbo

instances of the word "reporters" along with associated dialogue or quotes

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

tein

-0.78

ez

-0.70

uras

-0.69

inho

-0.69

ham

-0.66

amic

-0.65

ston

-0.65

vasive

-0.64

cer

-0.63

emonic

-0.63

POSITIVE LOGITS

 afterward

0.83

 conference

0.82

 gathered

0.82

 reporters

0.80

 assembled

0.75

peak

0.74

 onstage

0.74

 briefing

0.73

hips

0.73

 corps

0.73

Activations Density 0.030%

occurrences related to statements made by individuals to the press

instances of the word "reporters" along with associated dialogue or quotes

No Comments

No Known Activations