INDEX

Explanations

websites related to WikiLeaks

oai_token-act-pair · gpt-3.5-turbo

references to "Wikileaks" and related terms

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

Not in Any Lists

No Comments

Negative Logits

 mamm

-0.66

↵Âł

-0.64

 duplication

-0.62

 onwards

-0.61

 Barnett

-0.60

 Drift

-0.59

 Sagan

-0.58

 Author

-0.58

 Ferr

-0.58

 Patient

-0.57

POSITIVE LOGITS

kees

1.07

rants

1.03

keye

1.02

kered

0.95

estone

0.93

kee

0.90

untu

0.87

isoft

0.87

ety

0.85

keyes

0.83

Activations Density 0.054%

websites related to WikiLeaks

references to "Wikileaks" and related terms

No Comments

No Known Activations