INDEX

Explanations

single uppercase letters or numbers that could represent different points or locations

oai_token-act-pair · gpt-3.5-turbo

references to specific locations or points in a structured context, such as "point A" and "point B."

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

blogspot

-0.61

 toget

-0.61

season

-0.61

anan

-0.61

âĹ¼

-0.60

 twitch

-0.60

OPS

-0.59

ann

-0.58

 Paran

-0.57

fax

-0.57

POSITIVE LOGITS

cca

0.76

gregation

0.74

rimination

0.73

uties

0.70

ourke

0.69

arten

0.66

older

0.64

æ©

0.63

 dstg

0.62

ciation

0.61

Activations Density 0.214%

single uppercase letters or numbers that could represent different points or locations

references to specific locations or points in a structured context, such as "point A" and "point B."

No Comments

No Known Activations