INDEX

Explanations

links or prompts for user input

oai_token-act-pair · gpt-3.5-turbo

instances where the word "Enter" appears, which is commonly associated with prompts or instructions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Archangel

-0.71

 backdrop

-0.69

 conviction

-0.67

 Panther

-0.61

 asset

-0.59

 Bulletin

-0.58

packing

-0.58

 laundering

-0.58

 goodwill

-0.57

 Bastard

-0.57

POSITIVE LOGITS

prise

1.68

prises

1.62

tainment

1.56

prising

1.41

tain

1.33

taining

1.31

tained

0.94

ipher

0.86

prisingly

0.85

ymes

0.84

Activations Density 0.017%

links or prompts for user input

instances where the word "Enter" appears, which is commonly associated with prompts or instructions

No Comments

No Known Activations