INDEX

Explanations

phrases or words related to drawing attention to specific topics or issues

oai_token-act-pair · gpt-3.5-turbo

phrases related to drawing or attracting attention

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

tein

-0.75

 halves

-0.75

iche

-0.64

 Recon

-0.61

Sic

-0.61

 Pist

-0.60

 Tale

-0.60

 Yugoslavia

-0.58

 Whites

-0.58

raz

-0.57

POSITIVE LOGITS

estinal

0.84

orial

0.82

 seeker

0.77

 largeDownload

0.75

 attention

0.75

 seekers

0.74

stadt

0.73

arios

0.73

icter

0.73

 Attention

0.70

Activations Density 0.014%

phrases or words related to drawing attention to specific topics or issues

phrases related to drawing or attracting attention

No Comments

No Known Activations