INDEX

Explanations

statements emphasizing an unexpected or extreme outcome or circumstance

oai_token-act-pair · gpt-3.5-turbo

instances of the word "even."

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

rend

-0.74

idon

-0.68

alogy

-0.64

_-

-0.63

plex

-0.63

eme

-0.62

ugal

-0.60

=-=-=-=-=-=-=-=-

-0.60

ymph

-0.60

rog

-0.60

POSITIVE LOGITS

 remotely

1.13

 bother

0.78

 mention

0.76

 bothered

0.75

 bothering

0.74

 slightest

0.71

 mentioning

0.70

 outright

0.69

THING

0.66

handed

0.66

Activations Density 0.033%

statements emphasizing an unexpected or extreme outcome or circumstance

instances of the word "even."

No Comments

No Known Activations