INDEX

Explanations

words related to physical objects or locations within a community

oai_token-act-pair · gpt-3.5-turbo

words related to food and cooking

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

IJ

-0.71

 Sunshine

-0.69

QC

-0.61

enhagen

-0.60

Kes

-0.60

Neb

-0.60

 Scand

-0.59

Nem

-0.59

ONSORED

-0.59

RAG

-0.59

POSITIVE LOGITS

omial

0.78

boarding

0.77

dropping

0.76

ucker

0.75

matic

0.74

hole

0.74

dust

0.74

geoning

0.73

 coats

0.73

onomic

0.72

Activations Density 0.349%

words related to physical objects or locations within a community

words related to food and cooking

No Comments

No Known Activations