INDEX

Explanations

references to sports events, specifically related to kickoffs and Kickstarters

oai_token-act-pair · gpt-3.5-turbo

references to football-related terms, particularly the word "kick."

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

 abund

-0.84

BuyableInstoreAndOnline

-0.74

 Emin

-0.74

Dialogue

-0.73

icity

-0.69

afety

-0.68

 Debor

-0.68

å¤©

-0.67

Else

-0.67

icol

-0.66

POSITIVE LOGITS

starter

1.17

boxing

1.05

strap

1.01

starting

0.98

started

0.94

backs

0.86

haw

0.85

outs

0.83

start

0.82

ayne

0.81

Activations Density 0.018%

references to sports events, specifically related to kickoffs and Kickstarters

references to football-related terms, particularly the word "kick."

No Comments

No Known Activations