INDEX

Explanations

mentions of people named Jeremy

oai_token-act-pair · gpt-3.5-turbo

references to individuals named Jeremy, particularly in a political context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

holders

-0.76

maid

-0.75

liness

-0.74

ship

-0.73

ties

-0.71

ships

-0.69

ledged

-0.66

hips

-0.65

 understatement

-0.65

ional

-0.63

POSITIVE LOGITS

 Corbyn

1.30

 Clarkson

1.01

Pax

0.90

 Hammond

0.87

 Bent

0.86

 Hunt

0.84

urger

0.81

IELD

0.77

 Vine

0.76

 Buckingham

0.75

Activations Density 0.043%

mentions of people named Jeremy

references to individuals named Jeremy, particularly in a political context

No Comments

No Known Activations