INDEX

Explanations

mentions of violence and destruction

oai_token-act-pair · gpt-3.5-turbo

phrases related to violent actions and destruction

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

wcsstore

-0.63

 Canaver

-0.61

 prag

-0.58

 Gaal

-0.58

 Lessons

-0.57

 Athe

-0.57

 hindsight

-0.57

 Announce

-0.56

 Unsure

-0.56

APP

-0.56

POSITIVE LOGITS

ãĥł

0.72

 unprotected

0.67

 innocent

0.67

 throats

0.64

 entire

0.63

 livelihood

0.62

 innoc

0.60

 inhabitants

0.60

aneously

0.60

 pulp

0.59

Activations Density 0.850%

mentions of violence and destruction

phrases related to violent actions and destruction

No Comments

No Known Activations