INDEX

Explanations

terms related to individuals, specifically focusing on titles like "man" and "person."

oai_token-act-pair · gpt-3.5-turbo

references to individuals, particularly males

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

iolet

-0.66

 Advertisement

-0.63

arna

-0.60

 Pact

-0.59

iterranean

-0.59

rontal

-0.59

ominium

-0.57

erald

-0.57

 Despair

-0.56

illus

-0.56

POSITIVE LOGITS

 closest

1.14

liest

1.06

 nearest

1.06

 responsible

1.04

iest

0.90

responsible

0.89

 furthe

0.85

 pictured

0.84

who

0.83

osphere

0.82

Activations Density 0.279%

terms related to individuals, specifically focusing on titles like "man" and "person."

references to individuals, particularly males

No Comments

No Known Activations