INDEX

Explanations

phrases related to completion or closure

oai_token-act-pair · gpt-3.5-turbo

conclusive statements or assertions in a text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 cleans

-0.70

 exemplary

-0.69

quit

-0.69

 profits

-0.69

 lifes

-0.69

 salute

-0.68

escent

-0.66

 rejuven

-0.66

enture

-0.65

itar

-0.63

POSITIVE LOGITS

 Either

1.08

 Presumably

1.04

 Possibly

1.02

Either

0.96

Rum

0.93

 Probably

0.91

 Maybe

0.90

 Unsure

0.89

 Perhaps

0.89

 Suff

0.88

Activations Density 0.579%

phrases related to completion or closure

conclusive statements or assertions in a text

No Comments

No Known Activations