INDEX

Explanations

specific locations and incidents related to violent events

oai_token-act-pair · gpt-3.5-turbo

references to incidents of violence or attacks against individuals or groups

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

rogram

-0.80

Marsh

-0.68

 evolving

-0.66

 Siber

-0.65

Proof

-0.61

 defin

-0.59

ONEY

-0.58

 eminent

-0.58

 States

-0.58

Phys

-0.57

POSITIVE LOGITS

 belonging

0.78

 premises

0.73

 hostage

0.72

esson

0.72

 Annex

0.71

 hostages

0.70

 convoy

0.70

 buildings

0.69

area

0.68

 victim

0.66

Activations Density 0.530%

specific locations and incidents related to violent events

references to incidents of violence or attacks against individuals or groups

No Comments

No Known Activations