INDEX

Explanations

names of various individuals, likely with a focus on capturing first names

oai_token-act-pair · gpt-3.5-turbo

references to names and initials, particularly in the context of people

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 Romanian

-0.75

 Nicarag

-0.64

Ö¼

-0.61

 genocide

-0.61

 Croatian

-0.61

âĶĢâĶĢ

-0.61

ruary

-0.61

UA

-0.59

 Kenyan

-0.58

 Ethiopian

-0.58

POSITIVE LOGITS

antz

0.84

lake

0.81

Ĥ¬

0.75

otta

0.73

acco

0.73

itz

0.72

aghan

0.72

fork

0.70

kov

0.69

oshenko

0.69

Activations Density 0.123%

names of various individuals, likely with a focus on capturing first names

references to names and initials, particularly in the context of people

No Comments

No Known Activations