INDEX

Explanations

references to locations or entities related to the name "Wood"

oai_token-act-pair · gpt-3.5-turbo

references to the term "Woodrow Wilson"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

ional

-0.75

uates

-0.73

 disenfranch

-0.68

 disguise

-0.67

ysis

-0.67

uate

-0.66

itia

-0.66

 aver

-0.65

 analges

-0.65

uated

-0.64

POSITIVE LOGITS

ruff

1.46

working

1.17

stock

1.16

lands

1.11

lake

1.09

bridge

1.09

leigh

1.06

row

1.06

cutting

1.06

cock

1.05

Activations Density 0.033%

references to locations or entities related to the name "Wood"

references to the term "Woodrow Wilson"

No Comments

No Known Activations