INDEX

Explanations

mentions of underground tunnels

oai_token-act-pair · gpt-3.5-turbo

references to tunnels

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

itia

-0.76

======

-0.69

 Liberties

-0.66

EV

-0.66

 Summers

-0.66

ODUCT

-0.66

orian

-0.66

================================================================

-0.64

ģĸ

-0.64

CHO

-0.62

POSITIVE LOGITS

 tunnel

1.27

 tunnels

1.17

 boring

0.88

 Tunnel

0.88

dug

0.86

 handshake

0.85

 bunker

0.84

 entrances

0.82

ascus

0.81

 drilled

0.80

Activations Density 0.007%

mentions of underground tunnels

references to tunnels

No Comments

No Known Activations