INDEX

Explanations

phrases indicating a decision or conclusion

oai_token-act-pair · gpt-3.5-turbo

instances of the word "ruled" or its variations, often in a legal or declarative context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

issance

-0.85

illin

-0.76

velength

-0.74

akra

-0.72

Newsletter

-0.72

ufact

-0.70

ription

-0.70

ionage

-0.68

ickr

-0.68

 sqor

-0.65

POSITIVE LOGITS

phas

0.80

ihad

0.70

inker

0.69

maker

0.69

 unfit

0.67

uled

0.65

 against

0.65

 differently

0.64

 decisively

0.64

out

0.63

Activations Density 0.024%

phrases indicating a decision or conclusion

instances of the word "ruled" or its variations, often in a legal or declarative context

No Comments

No Known Activations