INDEX

Explanations

words related to comparison or composition

oai_token-act-pair · gpt-3.5-turbo

instances of the word "comp" and its derivatives, indicating a focus on compulsive behaviors and related terminology

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

Nig

-0.71

 Valencia

-0.68

hover

-0.67

Alm

-0.66

 Elves

-0.65

 Lans

-0.65

 Barcl

-0.65

 Borough

-0.64

 McLaren

-0.62

 Rite

-0.62

POSITIVE LOGITS

endium

1.21

 comp

1.18

ulsive

1.07

ilers

1.06

artments

1.06

ensation

0.98

rehens

0.95

rehensive

0.91

ressing

0.88

etitive

0.87

Activations Density 0.005%

words related to comparison or composition

instances of the word "comp" and its derivatives, indicating a focus on compulsive behaviors and related terminology

No Comments

No Known Activations