INDEX

Explanations

specific phrases including the word "hem"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "hem" and its related variations

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

âķĲ

-0.75

çĶŁ

-0.74

AVG

-0.74

ULTS

-0.72

 Peaks

-0.70

Oro

-0.69

raught

-0.64

 Puerto

-0.63

ÄŁ

-0.62

 Saras

-0.61

POSITIVE LOGITS

isphere

1.16

oglobin

1.11

oths

1.08

icals

0.99

med

0.99

ophon

0.95

ming

0.93

selves

0.90

icol

0.89

useum

0.87

Activations Density 0.005%

specific phrases including the word "hem"

occurrences of the word "hem" and its related variations

No Comments

No Known Activations