INDEX

Explanations

phrases that indicate consideration or taking something into account

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate consideration or assessment of various factors

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

sil

-0.77

quer

-0.76

si

-0.73

cop

-0.72

numbered

-0.71

edition

-0.70

bec

-0.70

code

-0.69

codes

-0.69

fruit

-0.68

POSITIVE LOGITS

enance

1.00

 GOODMAN

0.80

 Dame

0.73

 deductions

0.71

 factors

0.69

 respecting

0.66

agall

0.66

how

0.64

 inputs

0.64

 Dragonbound

0.64

Activations Density 0.018%

phrases that indicate consideration or taking something into account

phrases that indicate consideration or assessment of various factors

No Comments

No Known Activations