INDEX

Explanations

personal achievements or accolades in a sports context

oai_token-act-pair · gpt-3.5-turbo

phrases indicating player performance and statistics in sports

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

orno

-0.76

 Straw

-0.73

 Norn

-0.71

wives

-0.70

 Settlement

-0.70

sonian

-0.70

 Parties

-0.70

brow

-0.69

 Membership

-0.69

ESE

-0.69

POSITIVE LOGITS

 undrafted

0.93

 rehab

0.89

 relieved

0.88

 preseason

0.87

 injury

0.85

 infield

0.85

 batted

0.85

 athlet

0.85

 halftime

0.84

 clutch

0.84

Activations Density 0.391%

personal achievements or accolades in a sports context

phrases indicating player performance and statistics in sports

No Comments

No Known Activations