INDEX

Explanations

terms related to software commands and instructions

oai_token-act-pair · gpt-3.5-turbo

future tense verbs referring to actions or processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

soType

-0.86

 mastering

-0.68

soever

-0.67

 Flavoring

-0.66

Said

-0.66

Sean

-0.66

 vocational

-0.65

quished

-0.65

 Lawrence

-0.64

 departures

-0.64

POSITIVE LOGITS

 automatically

1.24

 overwrite

1.17

 expire

1.11

 appear

1.08

 popup

1.07

 notify

1.05

 emit

1.03

 prompt

1.02

 disappear

1.01

 generate

1.00

Activations Density 0.145%

terms related to software commands and instructions

future tense verbs referring to actions or processes

No Comments

No Known Activations