INDEX

Explanations

words related to specific named entities or terms, such as "Chipotle", "biotin", and "bigot"

oai_token-act-pair · gpt-3.5-turbo

words related to specific types of food or dietary components

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 confir

-0.69

 exha

-0.63

 desperate

-0.63

nesota

-0.62

 continuum

-0.61

inges

-0.60

vow

-0.59

 cohorts

-0.59

 sentence

-0.59

 thirst

-0.59

POSITIVE LOGITS

cot

1.42

chio

1.23

arget

1.06

otle

1.02

olkien

0.97

rans

0.94

pole

0.92

uning

0.91

agon

0.91

oxin

0.89

Activations Density 0.006%

words related to specific named entities or terms, such as "Chipotle", "biotin", and "bigot"

words related to specific types of food or dietary components

No Comments

No Known Activations