INDEX

Explanations

phrases related to initiating actions or activities

oai_token-act-pair · gpt-3.5-turbo

instances of the word "set" 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 @ 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

ability

-0.73

vernment

-0.69

lez

-0.69

ividual

-0.66

maxwell

-0.62

fighting

-0.61

 frivol

-0.61

ertodd

-0.61

rieved

-0.61

alez

-0.60

POSITIVE LOGITS

tle

1.39

 aside

1.23

 sail

1.22

abl

1.08

 forth

1.04

 sights

0.96

up

0.93

 foot

0.87

 alarms

0.84

 fires

0.84

Activations Density 0.031%

phrases related to initiating actions or activities

instances of the word "set" in various contexts

No Comments

No Known Activations