INDEX

Explanations

words related to historical locations or people

oai_token-act-pair · gpt-3.5-turbo

words related to children or childhood

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

andise

-0.98

okin

-0.76

 Dragonbound

-0.71

 Curry

-0.68

éŃĶ

-0.67

ADE

-0.67

ãĤ®

-0.67

oleon

-0.66

perature

-0.66

displayText

-0.65

POSITIVE LOGITS

erers

0.91

reth

0.87

rag

0.80

doms

0.80

er

0.80

sburg

0.78

0.77

itudinal

0.75

erer

0.74

roid

0.73

Activations Density 0.027%

words related to historical locations or people

words related to children or childhood

No Comments

No Known Activations