INDEX

Explanations

terms related to political figures such as lawmakers, legislators, MPs, and congressmen

oai_token-act-pair · gpt-3.5-turbo

references to political figures and lawmakers

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

uras

-0.72

 conquest

-0.64

phal

-0.64

 caliphate

-0.63

istic

-0.63

 conqu

-0.60

 Prophe

-0.60

OUS

-0.60

tern

-0.60

clusive

-0.59

POSITIVE LOGITS

hips

1.03

who

0.98

 sponsoring

0.90

crit

0.88

hip

0.88

resent

0.86

ervatives

0.86

 representing

0.85

 Dianne

0.83

woman

0.82

Activations Density 0.086%

terms related to political figures such as lawmakers, legislators, MPs, and congressmen

references to political figures and lawmakers

No Comments

No Known Activations