INDEX

Explanations

names of public figures or professions, with particular emphasis on comedians

oai_token-act-pair · gpt-3.5-turbo

mentions of comedians and related performers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

 Pradesh

-0.84

 Territories

-0.74

ories

-0.67

enza

-0.66

doors

-0.64

 differential

-0.62

 Matter

-0.62

plings

-0.61

topic

-0.60

eanor

-0.60

POSITIVE LOGITS

ess

0.96

 laureate

0.96

who

0.95

esses

0.94

 extraord

0.88

who

0.78

pocket

0.76

 beware

0.68

 whose

0.67

Artist

0.67

Activations Density 0.074%

names of public figures or professions, with particular emphasis on comedians

mentions of comedians and related performers

No Comments

No Known Activations