INDEX

Explanations

commands or prompts related to software or technology operations

oai_token-act-pair · gpt-3.5-turbo

commands or instructions related to unlocking features or access

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

itcher

-0.69

Mobile

-0.66

beer

-0.65

 Variant

-0.64

 Pend

-0.62

APD

-0.62

 Fenrir

-0.62

ãĥ´ãĤ¡

-0.61

mc

-0.60

USL

-0.58

POSITIVE LOGITS

 Unlock

0.74

ntil

0.71

 commencement

0.70

 Letters

0.68

ihad

0.66

assador

0.66

=-=-=-=-

0.64

ailability

0.63

ermanent

0.62

acca

0.61

Activations Density 0.000%

commands or prompts related to software or technology operations

commands or instructions related to unlocking features or access

No Comments

No Known Activations