INDEX

Explanations

proper nouns, particularly names of individuals

oai_token-act-pair · gpt-3.5-turbo

references to the name "Ger" or related names 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 @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

netflix

-0.82

uncture

-0.81

essee

-0.75

unct

-0.71

SPONSORED

-0.70

wikipedia

-0.70

yright

-0.67

piracy

-0.64

intendent

-0.64

plain

-0.64

POSITIVE LOGITS

Ger

1.13

deen

0.88

anyl

0.87

anium

0.85

rera

0.83

 warr

0.79

 resil

0.78

sten

0.77

ger

0.77

alf

0.76

Activations Density 0.003%

proper nouns, particularly names of individuals

references to the name "Ger" or related names in various contexts

No Comments

No Known Activations