INDEX

Explanations

words related to soccer teams, players, and matches

oai_token-act-pair · gpt-3.5-turbo

references to specific football clubs, particularly Manchester City

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

orable

-0.77

Nek

-0.75

 Hawaiian

-0.69

Narr

-0.69

Larry

-0.69

insula

-0.68

bip

-0.68

achusetts

-0.66

 Scal

-0.65

 Cindy

-0.65

POSITIVE LOGITS

 footballer

1.01

 Football

0.96

 striker

0.94

 midfielder

0.94

liga

0.93

 Bagg

0.92

 goalkeeper

0.92

 Tottenham

0.88

 Rugby

0.86

 Sunderland

0.82

Activations Density 0.174%

words related to soccer teams, players, and matches

references to specific football clubs, particularly Manchester City

No Comments

No Known Activations