INDEX

Explanations

descriptions of criminal activities or events involving law enforcement

oai_token-act-pair · gpt-3.5-turbo

commas and patterns of incomplete sentences in reporting incidents

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

Negative Logits

oliberal

-0.73

NetMessage

-0.69

uble

-0.69

agine

-0.68

ophon

-0.67

eem

-0.67

ocl

-0.67

Quantity

-0.66

 utopian

-0.66

leground

-0.66

POSITIVE LOGITS

 according

1.11

 prompting

0.91

TMZ

0.91

 authorities

0.88

 prosecutors

0.87

 injuring

0.85

 spokeswoman

0.85

according

0.84

 paramedics

0.82

 coroner

0.82

Activations Density 0.330%

descriptions of criminal activities or events involving law enforcement

commas and patterns of incomplete sentences in reporting incidents

No Known Activations