INDEX

Explanations

phrases related to beginnings or initial occurrences

oai_token-act-pair · gpt-3.5-turbo

references to significant first occurrences or events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 unfocusedRange

-0.73

nature

-0.68

utils

-0.68

nit

-0.68

 Powers

-0.67

ensitivity

-0.66

 abound

-0.65

affer

-0.65

Gameplay

-0.63

cffff

-0.63

POSITIVE LOGITS

imester

0.95

 imaginable

0.86

 installment

0.85

 iteration

0.85

 incarnation

0.84

 batch

0.80

 hurdle

0.77

 sighting

0.77

 outing

0.76

 ever

0.76

Activations Density 0.334%

phrases related to beginnings or initial occurrences

references to significant first occurrences or events

No Comments

No Known Activations