INDEX

Explanations

statements or questions expressing disbelief or incredulity

oai_token-act-pair · gpt-3.5-turbo

expressions of disbelief or incredulity

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

catentry

-0.94

lez

-0.76

Runner

-0.70

ipple

-0.64

 Ripple

-0.63

 skeleton

-0.62

 Survivors

-0.61

ateral

-0.61

 Graveyard

-0.60

aina

-0.59

POSITIVE LOGITS

 suddenly

0.80

ammed

0.80

DonaldTrump

0.74

anything

0.73

 knowingly

0.73

 anyone

0.72

 blindly

0.71

 anybody

0.70

 otherwise

0.69

 bothered

0.68

Activations Density 0.249%

statements or questions expressing disbelief or incredulity

expressions of disbelief or incredulity

No Comments

No Known Activations