INDEX

Explanations

contractions of "didn't."

oai_token-act-pair · gpt-3.5-turbo

instances of the word "didn't."

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

 Dise

-0.75

upp

-0.68

armed

-0.68

 DRAGON

-0.66

mania

-0.65

 Butt

-0.64

 Duty

-0.62

 Techniques

-0.61

ugu

-0.61

dar

-0.60

POSITIVE LOGITS

't

1.24

etsk

0.93

nt

0.77

kered

0.76

geon

0.76

okia

0.76

nels

0.75

ÃŃ

0.74

ned

0.71

":"/

0.69

Activations Density 0.060%

contractions of "didn't."

instances of the word "didn't."

No Comments

No Known Activations