INDEX

Explanations

words related to accusations or claims

oai_token-act-pair · gpt-3.5-turbo

repeated references to allegations and claims regarding misconduct or wrongdoing

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

Not in Any Lists

No Comments

Negative Logits

patch

-0.77

OVA

-0.73

otle

-0.72

ARCH

-0.72

 Britann

-0.70

avascript

-0.70

toggle

-0.69

ularity

-0.69

lake

-0.66

reen

-0.65

POSITIVE LOGITS

 violations

0.85

 abuser

0.83

 wrongdoing

0.83

 extortion

0.83

 abuses

0.83

 wrongful

0.83

 abuse

0.82

 abusers

0.82

 allegations

0.82

 alleges

0.81

Activations Density 0.051%

words related to accusations or claims

repeated references to allegations and claims regarding misconduct or wrongdoing

No Comments

No Known Activations