INDEX

Explanations

phrases related to inquiries about people or identity

oai_token-act-pair · gpt-3.5-turbo

inquiries about identity and authorship 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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

unctions

-0.72

umsy

-0.72

ritical

-0.70

taboola

-0.69

avery

-0.67

complex

-0.65

ecause

-0.65

urances

-0.65

osures

-0.65

recy

-0.64

POSITIVE LOGITS

liest

0.76

 hardest

0.74

 whom

0.73

?",

0.72

?),

0.71

iest

0.71

?,

0.66

0.64

 litter

0.63

?"

0.62

Activations Density 0.446%

phrases related to inquiries about people or identity

inquiries about identity and authorship in various contexts

No Comments

No Known Activations