INDEX

Explanations

tags or labels in text

oai_token-act-pair · gpt-3.5-turbo

mentions of the word "tag."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

theless

-0.85

ITNESS

-0.74

 Lumpur

-0.72

 Reverend

-0.70

¬¼

-0.68

 Seym

-0.67

Cox

-0.67

etheless

-0.63

 Scand

-0.63

Ell

-0.62

POSITIVE LOGITS

alog

0.98

gers

0.98

tags

0.97

tag

0.97

 tags

0.93

tag

0.92

ged

0.90

gery

0.88

otle

0.83

masters

0.82

Activations Density 0.008%

tags or labels in text

mentions of the word "tag."

No Comments

No Known Activations