INDEX

Explanations

references to various criminal accusations

oai_token-act-pair · gpt-3.5-turbo

instances of the word "accused" related to various crimes and allegations

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

ARCH

-0.78

emetery

-0.77

ental

-0.72

ooked

-0.70

ables

-0.69

resent

-0.68

Exit

-0.67

ahime

-0.67

mentation

-0.66

patch

-0.66

POSITIVE LOGITS

onia

0.79

 accused

0.78

isance

0.77

 accuse

0.75

 accuses

0.75

 accusing

0.73

heet

0.73

 wrongly

0.72

 Guilty

0.69

ilities

0.68

Activations Density 0.024%

references to various criminal accusations

instances of the word "accused" related to various crimes and allegations

No Known Activations