INDEX

Explanations

terms related to threats and violence

oai_token-act-pair · gpt-3.5-turbo

references to threats and violent actions

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

 Parables

-0.75

 guiName

-0.72

eret

-0.70

quickShipAvailable

-0.67

æ©Ł

-0.63

Reviewer

-0.63

aepernick

-0.63

 Costume

-0.63

ele

-0.63

 Videos

-0.62

POSITIVE LOGITS

 retaliation

1.07

 retribution

1.05

 repr

1.01

 eviction

0.95

 annihilation

0.93

 jeopard

0.84

 endanger

0.84

 blackmail

0.83

 wrath

0.82

 extinction

0.79

Activations Density 0.125%

terms related to threats and violence

references to threats and violent actions

No Comments

No Known Activations