INDEX

Explanations

Python syntax-related terms

oai_token-act-pair · gpt-3.5-turbo

references to syntax in programming and scripting languages

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

oppable

-0.83

 Kore

-0.82

zona

-0.75

 Gunn

-0.75

awei

-0.74

agher

-0.71

ECD

-0.70

 Gund

-0.69

atz

-0.69

 Fors

-0.67

POSITIVE LOGITS

 syntax

1.06

ntax

1.04

 highlighting

0.95

hess

0.86

ology

0.82

 translation

0.82

 keywords

0.81

 substitution

0.81

OLOGY

0.78

 parsing

0.78

Activations Density 0.020%

Python syntax-related terms

references to syntax in programming and scripting languages

No Comments

No Known Activations