INDEX

Explanations

causative relationships between actions and their effects or consequences

oai_token-act-pair · gpt-3.5-turbo

phrases that describe causes of problems or disruptions

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

apest

-0.79

Legend

-0.75

 congr

-0.73

arten

-0.71

soType

-0.69

aii

-0.69

worn

-0.69

ospels

-0.68

antage

-0.68

externalActionCode

-0.67

POSITIVE LOGITS

 ingest

0.92

 negligence

0.79

 ingestion

0.77

 displacement

0.75

 excessive

0.75

 neglect

0.74

 contamination

0.74

 Hurricane

0.73

dys

0.73

 Typhoon

0.73

Activations Density 0.206%

causative relationships between actions and their effects or consequences

phrases that describe causes of problems or disruptions

No Comments

No Known Activations