INDEX

Explanations

twists and unexpected developments in a narrative

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "twist" and its variations

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ufact

-0.74

leased

-0.68

pta

-0.66

 Ducks

-0.66

 appropriation

-0.66

rahim

-0.65

ogl

-0.65

league

-0.65

 Domain

-0.64

inez

-0.63

POSITIVE LOGITS

 twist

1.20

 twists

1.11

 Twist

0.89

 Whedon

0.82

 twisting

0.79

 weave

0.77

 angle

0.77

laus

0.74

hered

0.74

 stitch

0.73

Activations Density 0.031%

twists and unexpected developments in a narrative

occurrences of the word "twist" and its variations

No Comments

No Known Activations