INDEX

Explanations

names of individuals, likely celebrities or figures of public interest

oai_token-act-pair · gpt-3.5-turbo

proper nouns, particularly names of individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ashtra

-0.76

REDACTED

-0.75

incial

-0.72

âĶĢâĶĢ

-0.71

sonian

-0.70

 Pradesh

-0.69

iliate

-0.67

Italian

-0.67

ropolitan

-0.67

ensional

-0.67

POSITIVE LOGITS

iggs

0.73

lake

0.73

 Feather

0.71

 Howell

0.71

iffin

0.66

verson

0.66

monds

0.65

sworth

0.65

nington

0.65

houn

0.63

Activations Density 0.151%

names of individuals, likely celebrities or figures of public interest

proper nouns, particularly names of individuals

No Comments

No Known Activations