INDEX

Explanations

phrases related to technical documentation or instruction manuals

oai_token-act-pair · gpt-3.5-turbo

words related to instructions and guidelines

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

—-

-0.77

 somehow

-0.71

SPONSORED

-0.71

maybe

-0.70

whatever

-0.68

just

-0.66

.''

-0.65

nothing

-0.65

perhaps

-0.65

_.

-0.64

POSITIVE LOGITS

 calculator

0.86

 foregoing

0.83

 diagram

0.83

 Types

0.82

 guideline

0.75

 Approach

0.74

 Definitions

0.74

 Structure

0.74

 Finder

0.73

 Calculator

0.72

Activations Density 0.536%

phrases related to technical documentation or instruction manuals

words related to instructions and guidelines

No Comments

No Known Activations