INDEX

Explanations

words ending in specific patterns or a combination of letters and characters

oai_token-act-pair · gpt-3.5-turbo

words related to spelling and capitalization rules

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 Cunning

-0.71

Kin

-0.70

BuyableInstoreAndOnline

-0.69

Blu

-0.69

Temperature

-0.69

 Unlimited

-0.68

 overhe

-0.67

 Surviv

-0.67

geries

-0.67

 Creep

-0.67

POSITIVE LOGITS

 suffix

1.28

 prefix

1.16

 denote

1.11

 aster

1.10

 denotes

1.09

"-

1.08

"+

1.07

 brackets

1.06

 subscript

1.05

 notation

1.04

Activations Density 0.335%

words ending in specific patterns or a combination of letters and characters

words related to spelling and capitalization rules

No Comments

No Known Activations