INDEX

Explanations

spokeswomen mentioned in various news articles

oai_token-act-pair · gpt-3.5-turbo

references to spokespeople or public relations representatives

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

spons

-0.71

proc

-0.69

cru

-0.66

spir

-0.65

ech

-0.65

×

-0.64

inho

-0.64

ighth

-0.62

cens

-0.62

prop

-0.61

POSITIVE LOGITS

 Anne

0.84

arten

0.82

 Nicole

0.78

Anne

0.77

 spokeswoman

0.76

 Kathleen

0.74

 Gloria

0.72

 assistant

0.71

 Elise

0.71

ioned

0.70

Activations Density 0.017%

spokeswomen mentioned in various news articles

references to spokespeople or public relations representatives

No Comments

No Known Activations