INDEX

Explanations

adjectives related to importance or priority

oai_token-act-pair · gpt-3.5-turbo

keywords indicating the primary or dominant factors in various contexts

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

zanne

-0.84

oops

-0.82

earances

-0.77

ournals

-0.69

rentices

-0.69

rosso

-0.67

ldom

-0.67

rahim

-0.66

 Lauder

-0.66

Row

-0.66

POSITIVE LOGITS

 source

1.22

 beneficiary

1.19

 conduit

1.06

 culprit

1.04

 indicator

1.04

 obstacle

1.01

 predictor

0.99

 reason

0.99

 casualty

0.98

 contributor

0.96

Activations Density 0.130%

adjectives related to importance or priority

keywords indicating the primary or dominant factors in various contexts

No Comments

No Known Activations