INDEX

Explanations

mentions or references to engines and related components

oai_token-act-pair · gpt-3.5-turbo

references to engines

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

laus

-0.88

istani

-0.81

atana

-0.80

ties

-0.79

icho

-0.78

ities

-0.75

lyak

-0.75

 Jagu

-0.74

 Childhood

-0.72

onyms

-0.72

POSITIVE LOGITS

 powering

0.96

 compartment

0.85

 braking

0.84

bay

0.82

 combustion

0.79

 ignition

0.77

 horsepower

0.77

iHUD

0.76

ugal

0.75

 powered

0.75

Activations Density 0.046%

mentions or references to engines and related components

references to engines

No Comments

No Known Activations