INDEX

Explanations

the name "Morris" with varying levels of specificity

oai_token-act-pair · gpt-3.5-turbo

references to the name "Morris."

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

sheet

-0.76

fare

-0.67

mate

-0.67

sheets

-0.67

choes

-0.66

book

-0.65

REAM

-0.65

ccording

-0.64

 politic

-0.61

ware

-0.60

POSITIVE LOGITS

sey

1.77

ison

0.99

sein

0.92

olini

0.92

ons

0.91

seys

0.91

esi

0.91

hip

0.90

dale

0.87

beck

0.87

Activations Density 0.067%

the name "Morris" with varying levels of specificity

references to the name "Morris."

No Comments

No Known Activations