INDEX

Explanations

keywords related to the word "Arrow"

oai_token-act-pair · gpt-3.5-turbo

references to a specific TV show and its associated terms or entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

boarding

-0.74

boss

-0.69

 Rapt

-0.68

vill

-0.66

 Lara

-0.65

 Dion

-0.65

 starvation

-0.65

 quote

-0.65

Gor

-0.64

othy

-0.63

POSITIVE LOGITS

 Arrow

1.72

 renamed

1.52

ACE

1.48

 jointly

1.17

CW

1.11

IJ

1.10

LCS

1.07

 Flags

1.07

AMC

1.07

CW

1.05

Activations Density 0.052%

keywords related to the word "Arrow"

references to a specific TV show and its associated terms or entities

No Comments

No Known Activations