INDEX

Explanations

objects or entities located in close proximity to other objects or entities

oai_token-act-pair · gpt-3.5-turbo

references to proximity or adjacency in spatial contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

kay

-0.62

alth

-0.61

inger

-0.61

cks

-0.60

odium

-0.59

theless

-0.59

ismo

-0.59

hips

-0.58

olars

-0.58

krit

-0.57

POSITIVE LOGITS

 door

1.45

door

1.15

 Door

1.05

to

0.86

 doors

0.82

NetMessage

0.74

 aisle

0.73

 drawer

0.72

 doorway

0.70

 bend

0.68

Activations Density 0.023%

objects or entities located in close proximity to other objects or entities

references to proximity or adjacency in spatial contexts

No Comments

No Known Activations