INDEX

Explanations

references to a specific term, "Adele,"

oai_token-act-pair · gpt-3.5-turbo

the term "ele" in various contexts, often indicating an element or theme related to people or cultural events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

raint

-0.73

DPR

-0.71

ulate

-0.68

ossession

-0.68

raints

-0.67

 Kenobi

-0.67

ulum

-0.63

 Office

-0.63

ulated

-0.62

UAL

-0.61

POSITIVE LOGITS

phant

1.24

ttes

1.09

ghan

1.08

izabeth

1.02

venth

0.92

chy

0.86

ele

0.85

vez

0.84

fter

0.82

bians

0.81

Activations Density 0.011%

references to a specific term, "Adele,"

the term "ele" in various contexts, often indicating an element or theme related to people or cultural events

No Comments

No Known Activations