INDEX

Explanations

words related to physical string objects or concepts in different contexts

oai_token-act-pair · gpt-3.5-turbo

instances of the word "string."

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

tical

-0.85

icago

-0.76

hammad

-0.72

psey

-0.68

rentice

-0.68

psy

-0.65

hap

-0.65

egal

-0.65

ISSION

-0.65

mos

-0.64

POSITIVE LOGITS

ãĤ©

1.06

ãĤ¡

1.00

 string

0.96

enegger

0.86

ãĥĺ

0.85

 bikini

0.85

ãĥĥãĥī

0.81

 ItemLevel

0.78

 strings

0.78

string

0.78

Activations Density 0.007%

words related to physical string objects or concepts in different contexts

instances of the word "string."

No Comments

No Known Activations