INDEX

Explanations

names of individuals in positions of authority or leadership

oai_token-act-pair · gpt-3.5-turbo

names and titles related to law enforcement officials

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

âĹ¼

-0.75

 dstg

-0.70

 sidx

-0.69

Tokens

-0.68

Offline

-0.68

PLA

-0.68

VALUE

-0.67

 selfies

-0.66

 menstru

-0.66

pmwiki

-0.65

POSITIVE LOGITS

 Walsh

0.92

McC

0.89

 Burns

0.88

 Clayton

0.87

 Hutchinson

0.87

 Dixon

0.87

 Bennett

0.86

Zup

0.86

McL

0.86

 Nolan

0.86

Activations Density 0.188%

names of individuals in positions of authority or leadership

names and titles related to law enforcement officials

No Comments

No Known Activations