INDEX

Explanations

mathematical or programming-related expressions and formulas

oai_token-act-pair · gpt-3.5-turbo

phrases related to "for" in various contexts, often indicating purpose or function

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

Russ

-0.70

reat

-0.68

chenko

-0.67

qqa

-0.65

GREEN

-0.65

orah

-0.64

mare

-0.64

herty

-0.63

 embarrass

-0.63

ivil

-0.62

POSITIVE LOGITS

 example

1.14

 purposes

1.13

 instance

1.10

gery

1.06

bidden

1.02

 comparison

0.94

 sake

0.93

cing

0.90

 determining

0.89

 starters

0.89

Activations Density 0.253%

mathematical or programming-related expressions and formulas

phrases related to "for" in various contexts, often indicating purpose or function

No Comments

No Known Activations