INDEX

Explanations

capital letters followed by numbers that could potentially be part of an alphanumeric code

oai_token-act-pair · gpt-3.5-turbo

single-letter or abbreviated designations of organizations or entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

enegger

-1.04

shaw

-0.76

rador

-0.72

anamo

-0.71

ptin

-0.70

erella

-0.66

schild

-0.65

ãĥĥãĥī

-0.64

irement

-0.64

rawdownloadcloneembedreportprint

-0.62

POSITIVE LOGITS

ZE

0.85

TP

0.85

ND

0.84

OPS

0.83

Vs

0.83

GS

0.82

MS

0.81

ZI

0.80

USH

0.80

CE

0.80

Activations Density 0.080%

capital letters followed by numbers that could potentially be part of an alphanumeric code

single-letter or abbreviated designations of organizations or entities

No Comments

No Known Activations