INDEX

Explanations

mentions of mortars

oai_token-act-pair · gpt-3.5-turbo

terms related to construction or military weaponry, particularly various types of "mortar."

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

ãĥīãĥ©ãĤ´ãĥ³

-0.90

OPA

-0.82

 Ethiopia

-0.76

 Ethiopian

-0.75

 superpower

-0.74

 quotas

-0.72

BE

-0.72

DP

-0.66

 Zika

-0.66

 NAACP

-0.65

POSITIVE LOGITS

mort

3.93

 mort

3.64

 Mort

3.18

 mortar

1.90

mortem

1.47

 crem

1.18

emort

1.10

 Morty

1.10

 funer

1.04

 Morton

1.04

Activations Density 0.032%

mentions of mortars

terms related to construction or military weaponry, particularly various types of "mortar."

No Comments

No Known Activations