INDEX

Explanations

text containing the acronym "cs" with varying strengths of activation

oai_token-act-pair · gpt-3.5-turbo

instances of "cs" and "cb" related to computer science or coding terminology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 temptation

-0.72

 voting

-0.68

 conspicuous

-0.65

 redistributed

-0.64

 homophobic

-0.63

 publicity

-0.62

 JUSTICE

-0.61

 discrim

-0.61

ZIP

-0.61

 Mald

-0.61

POSITIVE LOGITS

cs

1.27

olars

1.00

vier

0.91

ules

0.87

ancel

0.87

uits

0.86

ouls

0.85

aucas

0.84

ubs

0.84

kies

0.84

Activations Density 0.006%

text containing the acronym "cs" with varying strengths of activation

instances of "cs" and "cb" related to computer science or coding terminology

No Comments

No Known Activations