INDEX

Explanations

phrases related to evidence or confirmation

oai_token-act-pair · gpt-3.5-turbo

the term "proven" in various contexts related to validation or evidence

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

adish

-0.75

letal

-0.72

ifle

-0.72

idays

-0.69

iewicz

-0.67

 squats

-0.66

eeper

-0.64

umbn

-0.64

paio

-0.64

onductor

-0.62

POSITIVE LOGITS

ãĥ¼ãĥĨ

0.97

 proven

0.92

iary

0.84

Ø³

0.80

ãĤ¤ãĥĪ

0.78

 refuted

0.78

ingen

0.78

Ô

0.75

 debunked

0.75

\\\\\\\\

0.74

Activations Density 0.015%

phrases related to evidence or confirmation

the term "proven" in various contexts related to validation or evidence

No Comments

No Known Activations