INDEX

Explanations

names related to legal investigations, especially the name "Mueller"

oai_token-act-pair · gpt-3.5-turbo

references to Robert Mueller and his investigative role

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

Pyr

-0.83

lihood

-0.70

 Barbarian

-0.69

points

-0.69

 Zoro

-0.69

 Fiji

-0.68

 Hawaiian

-0.67

draw

-0.66

 Columb

-0.65

piring

-0.64

POSITIVE LOGITS

 Mueller

0.99

 Probe

0.95

 probe

0.91

 indict

0.90

 investigation

0.88

 investigations

0.84

 probes

0.84

 probing

0.83

angelo

0.83

 investigating

0.82

Activations Density 0.023%

names related to legal investigations, especially the name "Mueller"

references to Robert Mueller and his investigative role

No Comments

No Known Activations