INDEX

Explanations

instances of the word "accept" and its variations in different contexts

oai_token-act-pair · gpt-3.5-turbo

forms of the word "accept."

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

borgh

-0.72

ionage

-0.70

geist

-0.70

iferation

-0.69

chn

-0.66

 glands

-0.66

borough

-0.66

xon

-0.66

 Indust

-0.65

brow

-0.65

POSITIVE LOGITS

 responsibility

0.98

ably

0.92

 bribes

0.83

ances

0.82

 accept

0.82

 acceptance

0.80

 submissions

0.79

 accepting

0.78

 defeat

0.77

ANCE

0.75

Activations Density 0.039%

instances of the word "accept" and its variations in different contexts

forms of the word "accept."

No Comments

No Known Activations