INDEX

Explanations

phrases related to the geographical location "Negev."

oai_token-act-pair · gpt-3.5-turbo

occurrences of the term "Ne" followed by other characters, suggesting a focus on references to neurons or related topics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

hips

-0.95

loo

-0.93

DOWN

-0.88

ãĥ¼ãĥĨãĤ£

-0.83

DragonMagazine

-0.83

MODE

-0.80

å§«

-0.74

REAM

-0.72

trap

-0.71

Ranked

-0.71

POSITIVE LOGITS

arest

1.22

cker

1.04

olithic

1.03

uron

1.02

oliberal

0.97

braska

0.96

verend

0.95

elin

0.93

umann

0.93

uve

0.90

Activations Density 0.008%

phrases related to the geographical location "Negev."

occurrences of the term "Ne" followed by other characters, suggesting a focus on references to neurons or related topics

No Comments

No Known Activations