INDEX

Explanations

names, specifically focusing on the name "Lamar"

oai_token-act-pair · gpt-3.5-turbo

references to the name "Lamar" in various contexts

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

IFrame

Link

Negative Logits

illing

-0.78

spring

-0.72

astical

-0.71

 Pension

-0.69

DIV

-0.69

sharp

-0.68

orks

-0.68

ħĭ

-0.66

OSP

-0.66

shock

-0.65

POSITIVE LOGITS

 Lamar

1.13

 Lyndon

0.77

gebra

0.76

ques

0.75

iens

0.75

 Ferr

0.71

 Marino

0.69

 Garr

0.69

inate

0.68

 Cain

0.68

Activations Density 0.018%

names, specifically focusing on the name "Lamar"

references to the name "Lamar" in various contexts

No Known Activations