INDEX

Explanations

instances of people being criticized or attacked for various reasons

oai_token-act-pair · gpt-3.5-turbo

the word "for" in various contexts, indicating a focus on prepositional phrases

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

illin

-0.82

edin

-0.72

atl

-0.71

Mine

-0.67

awan

-0.66

 Â®

-0.64

NET

-0.64

abo

-0.62

mare

-0.61

nan

-0.61

POSITIVE LOGITS

geries

1.10

 daring

1.03

 violating

1.02

 failing

1.00

 lack

0.95

 reasons

0.94

 refusing

0.93

 questioning

0.89

gery

0.88

 breaching

0.86

Activations Density 0.144%

instances of people being criticized or attacked for various reasons

the word "for" in various contexts, indicating a focus on prepositional phrases

No Comments

No Known Activations