INDEX

Explanations

future actions or decisions

oai_token-act-pair · gpt-3.5-turbo

instances of the word "do" and its variations, indicating inquiries or commands about actions

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

 Entered

-0.98

 Frie

-0.72

gart

-0.70

theless

-0.67

sent

-0.65

tro

-0.63

Ĭ±

-0.63

inently

-0.62

printed

-0.61

ware

-0.61

POSITIVE LOGITS

pez

1.09

atives

0.74

berman

0.70

etting

0.68

 wrong

0.66

ggy

0.65

INGS

0.65

 differently

0.65

":"

0.64

ients

0.63

Activations Density 0.063%

future actions or decisions

instances of the word "do" and its variations, indicating inquiries or commands about actions

No Comments

No Known Activations