INDEX

Explanations

words related to political figures and locations

oai_token-act-pair · gpt-3.5-turbo

names of individuals, particularly in the context of locations and titles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

+/-

-0.65

plays

-0.64

 SPACE

-0.63

 caps

-0.62

effic

-0.61

xp

-0.61

 diligence

-0.60

sense

-0.58

=-=-=-=-

-0.57

 culp

-0.57

POSITIVE LOGITS

adra

0.75

west

0.74

elaide

0.72

iste

0.71

HR

0.68

ascus

0.68

endment

0.67

ikh

0.66

riors

0.66

 Stores

0.66

Activations Density 0.348%

words related to political figures and locations

names of individuals, particularly in the context of locations and titles

No Comments

No Known Activations