INDEX

Explanations

names of military equipment or missions

oai_token-act-pair · gpt-3.5-turbo

references to video game titles and characters

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

emetery

-0.80

atural

-0.74

ormal

-0.73

onymous

-0.73

odes

-0.72

onym

-0.70

RPC

-0.69

ancial

-0.68

ograp

-0.68

osponsors

-0.66

POSITIVE LOGITS

Shock

0.83

shock

0.74

Jak

0.73

MK

0.73

heed

0.72

toe

0.70

MENTS

0.69

toe

0.68

Rog

0.67

Eye

0.66

Activations Density 0.021%

names of military equipment or missions

references to video game titles and characters

No Comments

No Known Activations