INDEX

Explanations

words related to police activities and investigations

oai_token-act-pair · gpt-3.5-turbo

words related to police actions and investigations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

isSpecialOrderable

-0.76

opian

-0.74

die

-0.72

 proportions

-0.70

 competitiveness

-0.65

 grapes

-0.65

ankind

-0.65

âķ

-0.64

Percent

-0.63

spir

-0.63

POSITIVE LOGITS

 Detect

0.97

 investigating

0.92

nsic

0.89

 paramedics

0.86

 snipers

0.86

nsics

0.86

 checkpoints

0.85

 patrols

0.83

 patrolling

0.83

 surveillance

0.82

Activations Density 0.462%

words related to police activities and investigations

words related to police actions and investigations

No Comments

No Known Activations