INDEX

Explanations

lines of code or code-related terms

oai_token-act-pair · gpt-3.5-turbo

mentions of "code" and related programming concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

 Flavoring

-1.01

iscopal

-0.81

iversal

-0.78

iatus

-0.77

 Wynne

-0.72

Ń·

-0.70

izo

-0.70

 Bever

-0.70

 Democr

-0.67

cffff

-0.65

POSITIVE LOGITS

 snippet

1.34

 snippets

1.15

base

1.10

code

0.88

 code

0.86

velop

0.83

 fragment

0.82

 repository

0.82

works

0.82

 snipp

0.81

Activations Density 0.025%

lines of code or code-related terms

mentions of "code" and related programming concepts

No Comments

No Known Activations