INDEX

Explanations

terms related to operational activities or functions

oai_token-act-pair · gpt-3.5-turbo

references to operational processes or activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hus

-0.77

arton

-0.72

sonian

-0.72

Ĥª

-0.72

apple

-0.69

birth

-0.66

inger

-0.65

marked

-0.64

aiman

-0.63

artz

-0.63

POSITIVE LOGITS

ally

1.04

 tempo

0.91

eering

0.87

 manual

0.82

 manuals

0.81

ality

0.81

 Manual

0.76

OPS

0.75

 conducted

0.72

SPONSORED

0.72

Activations Density 0.071%

terms related to operational activities or functions

references to operational processes or activities

No Comments

No Known Activations