INDEX

Explanations

names or references to the name "Jane"

oai_token-act-pair · gpt-3.5-turbo

references to the name "Jane" associated with various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

*/(

-0.89

orescent

-0.87

ctic

-0.83

idated

-0.80

cffff

-0.77

rophe

-0.76

PDATE

-0.75

ursed

-0.74

akespe

-0.74

rophic

-0.73

POSITIVE LOGITS

Doe

1.27

 Jane

0.98

 Aust

0.90

 Seymour

0.85

Jane

0.84

 Approximately

0.82

Roe

0.81

 Jacobs

0.79

ju

0.78

Ey

0.76

Activations Density 0.009%

names or references to the name "Jane"

references to the name "Jane" associated with various contexts

No Comments

No Known Activations