INDEX

Explanations

phrases related to distance or extreme positioning

oai_token-act-pair · gpt-3.5-turbo

phrases indicating distance or radical positions

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

advertisement

-0.67

 Temp

-0.67

itu

-0.61

oly

-0.61

iots

-0.61

urg

-0.60

iley

-0.59

ãĤ¡

-0.57

=#

-0.57

olas

-0.56

POSITIVE LOGITS

fetched

1.24

 away

1.12

 removed

1.10

 ahead

1.04

 gone

0.95

 along

0.91

awa

0.90

 apart

0.90

 behind

0.90

 distant

0.88

Activations Density 0.078%

phrases related to distance or extreme positioning

phrases indicating distance or radical positions

No Comments

No Known Activations