INDEX

Explanations

login-related terms and prompts

oai_token-act-pair · gpt-3.5-turbo

terms related to login and authentication processes

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

ahn

-0.79

×Ĳ

-0.72

×Ķ

-0.70

mol

-0.69

aper

-0.67

á½

-0.67

anus

-0.65

damage

-0.64

icz

-0.64

 Judah

-0.64

POSITIVE LOGITS

 login

1.06

 credentials

1.05

 Username

0.88

arantine

0.80

essed

0.78

itors

0.75

ername

0.75

 logger

0.74

istical

0.74

 username

0.74

Activations Density 0.015%

login-related terms and prompts

terms related to login and authentication processes

No Comments

No Known Activations