INDEX

Explanations

phrases describing actions or characteristics as positive or negative

oai_token-act-pair · gpt-3.5-turbo

articles and adjectives describing significant actions or attributes

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

mentioned

-0.88

coins

-0.75

bg

-0.74

msg

-0.73

items

-0.72

bots

-0.71

noon

-0.70

attr

-0.68

gi

-0.67

atin

-0.67

POSITIVE LOGITS

 precursor

0.91

 nuisance

0.90

versive

0.89

 legitimate

0.87

 roadmap

0.84

 contribution

0.83

 watershed

0.83

 threat

0.83

 continuation

0.82

 deviation

0.80

Activations Density 0.323%

phrases describing actions or characteristics as positive or negative

articles and adjectives describing significant actions or attributes

No Comments

No Known Activations