INDEX

Explanations

phrases indicating uncertainty or speculation

oai_token-act-pair · gpt-3.5-turbo

negative assessments or perceptions of various subjects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

pez

-0.93

instead

-0.78

 instead

-0.77

itton

-0.72

ilts

-0.69

 nonetheless

-0.68

 doubtless

-0.66

éĹĺ

-0.66

 undoubtedly

-0.64

 Rouge

-0.63

POSITIVE LOGITS

 anymore

1.26

 bothered

1.04

 remotely

1.00

nor

0.99

 anywhere

0.96

 anything

0.95

any

0.95

 necessarily

0.94

 whatsoever

0.94

 terribly

0.89

Activations Density 0.123%

phrases indicating uncertainty or speculation

negative assessments or perceptions of various subjects

No Comments

No Known Activations