INDEX

Explanations

mentions of shocking events or surprising occurrences

oai_token-act-pair · gpt-3.5-turbo

ideas and phrases that reflect significant global events or the reactions of people to those events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

ITNESS

-0.69

 accelerating

-0.64

 aborted

-0.62

 accelerated

-0.60

nant

-0.59

arate

-0.58

 interception

-0.57

 automation

-0.57

adobe

-0.56

ternity

-0.55

POSITIVE LOGITS

 itch

0.84

aughs

0.78

 sensibilities

0.76

 senses

0.70

 onlook

0.66

jah

0.66

eps

0.65

stakes

0.65

 heck

0.65

gyn

0.64

Activations Density 0.253%

mentions of shocking events or surprising occurrences

ideas and phrases that reflect significant global events or the reactions of people to those events

No Comments

No Known Activations