INDEX

Explanations

words or prefixes related to medical or technical procedures and descriptions

oai_token-act-pair · gpt-3.5-turbo

references to health-related concepts and conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

Reviewer

-0.94

 HELL

-0.68

NEC

-0.67

Omn

-0.65

 Symphony

-0.64

chwitz

-0.63

 partName

-0.63

UID

-0.62

HAM

-0.62

 Assass

-0.62

POSITIVE LOGITS

oes

0.86

opt

0.85

Ã©t

0.85

Ãª

0.84

age

0.83

arial

0.82

emale

0.81

adic

0.80

Georg

0.77

ette

0.77

Activations Density 0.638%

words or prefixes related to medical or technical procedures and descriptions

references to health-related concepts and conditions

No Comments

No Known Activations