INDEX

Explanations

violence-related incidents or crimes

oai_token-act-pair · gpt-3.5-turbo

occurrences of violent incidents and fatalities

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

Tokens

-0.67

 ç¥ŀ

-0.66

soDeliveryDate

-0.66

 ãĤµ

-0.65

çļ

-0.65

å¥³

-0.65

ãĤ´ãĥ³

-0.60

ahime

-0.60

isSpecialOrderable

-0.57

ankind

-0.57

POSITIVE LOGITS

 allegedly

0.86

 burglary

0.86

 burgl

0.81

 NYPD

0.80

 gunman

0.77

 police

0.76

 alleged

0.75

 LAPD

0.74

 abusive

0.74

 unidentified

0.73

Activations Density 1.506%

violence-related incidents or crimes

occurrences of violent incidents and fatalities

No Comments

No Known Activations