INDEX

Explanations

coordinates or instructions related to location and positioning

oai_token-act-pair · gpt-3.5-turbo

references to geographical or positional coordinates

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

anol

-0.83

tu

-0.73

ko

-0.71

onia

-0.71

question

-0.70

ammy

-0.69

susp

-0.67

deb

-0.67

omnia

-0.67

hillary

-0.66

POSITIVE LOGITS

 coordinates

1.40

 coordinate

1.11

inates

1.03

 coord

0.88

ãĥ¼ãĥĨ

0.79

 Coord

0.77

 Spac

0.76

ially

0.73

 tracts

0.72

 coordinating

0.72

Activations Density 0.019%

coordinates or instructions related to location and positioning

references to geographical or positional coordinates

No Comments

No Known Activations