INDEX

Explanations

strings containing words related to date, location, and formatting

oai_token-act-pair · gpt-3.5-turbo

the preposition "in" used frequently throughout the document

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

minist

-0.71

 ostensibly

-0.69

 supposedly

-0.64

lik

-0.63

 nailed

-0.61

 personally

-0.61

 offending

-0.60

 right

-0.60

 fortun

-0.59

POSITIVE LOGITS

clusions

0.93

clus

0.92

STRUCT

0.83

CLUS

0.82

 Conversation

0.81

activated

0.81

humans

0.79

initialized

0.78

 Brief

0.78

efficiency

0.78

Activations Density 0.042%

strings containing words related to date, location, and formatting

the preposition "in" used frequently throughout the document

No Comments

No Known Activations