INDEX

Explanations

phrases related to breaking or chipping something

oai_token-act-pair · gpt-3.5-turbo

words related to actions or processes of cutting, breaking, or altering states

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

rade

-0.69

ational

-0.62

cation

-0.62

izen

-0.61

 Siber

-0.58

she

-0.56

spe

-0.55

sat

-0.55

stad

-0.55

 Gret

-0.54

POSITIVE LOGITS

*/(

0.85

olicy

0.81

ules

0.81

utical

0.80

ilver

0.78

dates

0.77

rison

0.76

ule

0.75

 toes

0.74

hift

0.70

Activations Density 0.113%

phrases related to breaking or chipping something

words related to actions or processes of cutting, breaking, or altering states

No Comments

No Known Activations