INDEX

Explanations

sentiments of suspicion or controversy

oai_token-act-pair · gpt-3.5-turbo

terms related to accusations and unreliability in reporting

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

Negative Logits

ometers

-1.17

stories

-1.14

ernels

-1.06

iths

-1.02

forts

-1.01

pieces

-1.01

ships

-1.00

facts

-0.99

types

-0.98

events

-0.97

POSITIVE LOGITS

 version

0.97

 batch

0.96

 instance

0.92

 chunk

0.92

 portion

0.91

 allegation

0.90

 swath

0.89

 memorandum

0.88

 piece

0.87

 loophole

0.87

Activations Density 0.390%

sentiments of suspicion or controversy

terms related to accusations and unreliability in reporting

No Known Activations