INDEX

Explanations

information related to rankings or placements

oai_token-act-pair · gpt-3.5-turbo

references to rankings and positions in competitive contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

IRC

-0.69

 Patreon

-0.67

catentry

-0.65

embed

-0.61

ITNESS

-0.61

kens

-0.57

tumblr

-0.57

sav

-0.57

Learning

-0.56

soType

-0.56

POSITIVE LOGITS

 place

1.03

 overall

0.98

 nationally

0.89

 among

0.85

 Place

0.84

 amongst

0.83

irty

0.83

 baseman

0.75

 percentile

0.75

 placed

0.75

Activations Density 0.064%

information related to rankings or placements

references to rankings and positions in competitive contexts

No Comments

No Known Activations