INDEX

Explanations

mentions of situations involving peaks or high points

oai_token-act-pair · gpt-3.5-turbo

references to challenges or difficulties encountered in various contexts

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

 CODE

-0.65

TY

-0.64

 Reconstruction

-0.63

 Boxing

-0.63

 PUBLIC

-0.62

 Antar

-0.62

 Proposition

-0.61

 Taliban

-0.59

ãĥŁ

-0.59

 Ridley

-0.59

POSITIVE LOGITS

poons

1.30

etting

1.21

uits

1.18

hots

1.14

hip

1.11

etter

1.11

ensical

1.04

pace

1.04

cale

1.03

uggest

1.00

Activations Density 0.015%

mentions of situations involving peaks or high points

references to challenges or difficulties encountered in various contexts

No Comments

No Known Activations