INDEX

Explanations

names of political figures or officials

oai_token-act-pair · gpt-3.5-turbo

proper nouns, particularly names of people and political figures

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

geries

-0.78

age

-0.69

ĸļ

-0.67

 Dise

-0.66

 Reincarn

-0.65

 Experiment

-0.64

 Serve

-0.62

 mould

-0.61

Year

-0.61

 destiny

-0.61

POSITIVE LOGITS

 tweeted

1.54

 reacted

1.46

 applauded

1.44

 denounced

1.42

 congratulated

1.40

 lashed

1.32

 reiterated

1.31

 praised

1.31

 slammed

1.30

 condemned

1.30

Activations Density 0.394%

names of political figures or officials

proper nouns, particularly names of people and political figures

No Comments

No Known Activations