INDEX

Explanations

proper nouns or names, specifically "Keller"

oai_token-act-pair · gpt-3.5-turbo

mentions of specific names and organizations

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

ential

-0.90

Ö¼

-0.79

entials

-0.79

ENCE

-0.78

encia

-0.76

encies

-0.76

ences

-0.76

aud

-0.75

IBLE

-0.74

encers

-0.72

POSITIVE LOGITS

DPR

0.86

istani

0.81

Kat

0.79

kat

0.75

ratom

0.75

Kob

0.75

patrick

0.74

 Silk

0.73

wagen

0.73

 Helsinki

0.73

Activations Density 0.035%

proper nouns or names, specifically "Keller"

mentions of specific names and organizations

No Comments

No Known Activations