INDEX

Explanations

proper nouns, particularly names of people

oai_token-act-pair · gpt-3.5-turbo

instances of quoted speech or reported dialogue

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

clad

-0.69

luster

-0.64

oppable

-0.64

emed

-0.64

Gate

-0.60

Ale

-0.60

 advoc

-0.57

xtap

-0.56

 conflic

-0.55

cloth

-0.54

POSITIVE LOGITS

 msec

0.76

 goodbye

0.71

 href

0.65

 partName

0.63

 hello

0.62

reys

0.59

:]

0.58

=\"

0.57

0.54

 farewell

0.54

Activations Density 0.059%

proper nouns, particularly names of people

instances of quoted speech or reported dialogue

No Comments

No Known Activations