INDEX

Explanations

mentions of representatives or individuals representing something or someone

oai_token-act-pair · gpt-3.5-turbo

references to individuals or entities acting as representatives

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

frey

-0.97

aughs

-0.88

bows

-0.88

imb

-0.78

awar

-0.77

 Spit

-0.76

osure

-0.76

icals

-0.70

forth

-0.69

yard

-0.69

POSITIVE LOGITS

 representative

1.00

soType

0.80

 representatives

0.79

holder

0.75

 antioxid

0.72

 Representative

0.71

 Kislyak

0.70

ority

0.70

 android

0.70

 envoy

0.68

Activations Density 0.010%

mentions of representatives or individuals representing something or someone

references to individuals or entities acting as representatives

No Comments

No Known Activations