INDEX

Explanations

words related to legal terms or processes

oai_token-act-pair · gpt-3.5-turbo

suffixes that indicate medical or procedural terminology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

udeb

-0.78

yond

-0.65

herty

-0.65

Bey

-0.64

cmp

-0.63

izabeth

-0.61

ammad

-0.61

UTF

-0.61

SpaceEngineers

-0.61

itans

-0.60

POSITIVE LOGITS

 Advice

1.13

 Problems

1.11

 Overview

1.11

 Versus

1.11

 Issues

1.10

 Profile

1.10

 Comparison

1.08

 Removal

1.08

 Differences

1.08

 Benefits

1.08

Activations Density 0.346%

words related to legal terms or processes

suffixes that indicate medical or procedural terminology

No Comments

No Known Activations