INDEX

Explanations

names and positions of sports players, specifically focusing on the "forward" position

oai_token-act-pair · gpt-3.5-turbo

references to athletes playing as forwards in various sports

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

lez

-0.70

riz

-0.70

rip

-0.69

colo

-0.68

azo

-0.67

usting

-0.65

lement

-0.63

ripp

-0.63

gins

-0.63

 Pulitzer

-0.62

POSITIVE LOGITS

 thinker

0.78

Ħ¢

0.75

 forwards

0.70

ishop

0.69

srf

0.67

 Forward

0.66

forward

0.65

 pads

0.63

 puberty

0.63

userc

0.62

Activations Density 0.014%

names and positions of sports players, specifically focusing on the "forward" position

references to athletes playing as forwards in various sports

No Comments

No Known Activations